Kodesage, a Hungarian startup with registered offices in London that specialises in AI-powered solutions for legacy software modernisation, has announced pre-seed funding of \u00a31.9 million. <\/p>\n\n\n\n
PortfoLion Capital Partners, the VC and private equity arm of OTP Group, led the funding round. Angel investors, including three former UiPath senior executives and other enterprise software veterans, also participated. With existing paying customers and ongoing pilot programmes, the funding will be used to accelerate platform development and support the company\u2019s mission to help enterprises enhance their most complex legacy systems.<\/p>\n\n\n\n
Kodesage aims to address a critical issue in enterprise IT: understanding, maintaining and modernising legacy systems. These deeply embedded systems pose significant risks, including outdated technology, poor documentation, reliance on key employees and performance issues. <\/p>\n\n\n\n
Maintaining these systems often requires continuous updates, patches and ad hoc fixes from multiple sources, complicating employee training and increasing risks and costs. Market research highlights the urgency of this issue, with the global market for modernisation services projected to reach nearly \u00a360 billion in the next seven years. As organisations pursue digital transformation, the demand for AI-powered tools to streamline legacy system management and boost productivity has never been greater.<\/p>\n\n\n\n
Kodesage acts as a guide for engineers, simplifying code navigation and providing the support needed to upgrade even the most complex legacy systems without prior project knowledge. Built with enterprise in mind, the platform supports on-premise and even air-gapped deployments, ensuring compliance and data security. <\/p>\n\n\n\n
Technically, Kodesage\u2019s platform stands out due to its lower hardware requirements compared to solutions like OpenAI, enhancing cost efficiency. It is model agnostic, utilising the latest language learning models (LLMs) while allowing clients to choose their preferred models. Kodesage also generates a proprietary knowledge graph from various sources, including codebases, database schemas, past tests, issue tickets, wikis, and existing documentation. The firm\u2019s approach to language-specific parsing significantly improves LLM performance, particularly when dealing with complex legacy codebases.<\/p>\n\n\n\n